What is a typical permeability for clay?

Compacted lean clay (CL) runs about 1e-9 to 1e-7 cm/s and fat clay (CH) about 1e-10 to 1e-8 cm/s — practically impervious. For reference, a compacted clay liner is usually specified at 1e-7 cm/s or lower. Clean sand is six to nine orders of magnitude higher, roughly 1e-3 to 1e-1 cm/s.

How do you convert cm/s to ft/day?

Multiply cm/s by 2,835 to get ft/day. So 1e-3 cm/s = 2.84 ft/day. Other useful factors: 1 cm/s = 864 m/day = 21,200 gpd/sq ft = 1,417 in/hr, and 1 in/hr = 7.06e-4 cm/s.

What is Hazen's formula for permeability?

Hazen's empirical relation is k (cm/s) = C · D10², with D10 the effective grain size in millimeters and C between about 0.4 and 1.2 (1.0 is the common default). It applies only to clean, loose to medium-dense sands with a uniformity coefficient below about 5 and D10 between roughly 0.1 and 3.0 mm. It is a first estimate, not a design value — it can be off by an order of magnitude in real soils.

Which permeability test do I use for which soil?

Use the constant-head test (ASTM D2434) for coarse soils above about 1e-3 cm/s, the falling-head test for about 1e-3 to 1e-6 cm/s, and the flexible-wall (triaxial) permeameter (ASTM D5084) for fine-grained soils at or below 1e-6 cm/s. In the field use a pumping test or slug test below the water table, and a double-ring infiltrometer (ASTM D3385) or borehole permeameter above it.

What is soil permeability anisotropy?

Horizontal permeability usually exceeds vertical because of layering. Ratios of kh/kv of 2 to 10 are typical for natural stratified alluvium, 4 to 9 for compacted embankment fill built in lifts, and 10 to 100 or more for varved clays. Seepage analyses and flow nets are very sensitive to this ratio, so it should be assumed and stated explicitly rather than left at 1.

What Ksat defines the NRCS hydrologic soil groups?

NRCS defines the groups by the saturated hydraulic conductivity of the least transmissive layer: Group A above 5.67 in/hr (40 µm/s), Group B 1.42 to 5.67 in/hr (10 to 40 µm/s), Group C 0.14 to 1.42 in/hr (1.0 to 10 µm/s), and Group D below 0.14 in/hr (1.0 µm/s). Depth to a restrictive layer and to a seasonal high water table can override the conductivity band and force a dual group such as B/D.

All tools

Open embankment seepage calculator →

Soil Permeability (k) — Reference

By Michael Flynn, PE · water resources & dam safety engineer

Saturated hydraulic conductivity by soil type, the unit conversions that trip everyone up, and the test method that is actually valid at each magnitude. Permeability spans twelve orders of magnitude across ordinary soils — more than any other geotechnical parameter — so the useful question is almost never “what is k” but “what decade is k in, and how confident am I.”

Darcy's Law

v = k · i Q = k · i · A i = Δh / L
v_seepage = v / n (actual pore velocity; n = porosity)

v is the discharge velocity through the gross cross-section — it is not the speed of a water particle. Darcy's law holds for laminar flow, which covers essentially all soils finer than coarse gravel (Re < 1 based on D₁₀). In clean cobbles and rockfill it breaks down and a nonlinear Forchheimer form is needed.

Unit Conversions

From	→ cm/s	→ ft/day	→ m/day	→ in/hr	→ gpd/ft²
1 cm/s	1	2,835	864	1,417	21,200
1 ft/day	3.53×10⁻⁴	1	0.305	0.500	7.48
1 m/day	1.16×10⁻³	3.28	1	1.64	24.5
1 in/hr	7.06×10⁻⁴	2.00	0.610	1	15.0
1 µm/s	1×10⁻⁴	0.284	0.0864	0.142	2.12
1 gpd/ft²	4.72×10⁻⁵	0.134	0.0408	0.0669	1
1 darcy (water, 20°C)	≈9.6×10⁻⁴	2.7	0.83	1.4	20

Handy anchor: 1 cm/s ≈ 2,835 ft/day, and 1 in/hr ≈ 2 ft/day. NRCS soil surveys report µm/s; geotech reports report cm/s; groundwater models want ft/day; stormwater infiltration rules are written in in/hr.

Order of Magnitude & Drainage Class

k (cm/s)	Drainage	Typical soils	What it means in practice
10²–10⁰	Very good	Clean gravel, rockfill, open-graded stone	Free-draining; drain rock, chimney and blanket drains
10⁰–10⁻³	Good	Clean sands, clean sand-gravel mixtures	Drains under gravity; usable as a filter or infiltration receiver
10⁻³–10⁻⁵	Poor	Very fine sands, silts, silty/clayey sands, glacial till	Slow drainage; frost-susceptible; marginal for infiltration BMPs
10⁻⁵–10⁻⁷	Very poor	Silt, stratified clay, weathered clay fill	Effectively a barrier over construction time scales
< 10⁻⁷	Practically impervious	Homogeneous clays below the weathered zone, CCLs, GCLs	Liner and core material; 1×10⁻⁷ cm/s is the common CCL spec

Typical k by USCS Group (compacted fill)

USCS	k (cm/s)	k (ft/day)	Role in an embankment / earthwork
GW	10⁻²–10⁰	30–2,800	Pervious shell, drainage zone
GP	10⁻¹–10¹	300–28,000	Drain rock; needs a filter against migration
GM	10⁻⁶–10⁻³	0.003–3	Semi-pervious; usable core if fines are plastic
GC	10⁻⁸–10⁻⁶	3×10⁻⁵–0.003	Good core and blanket material
SW	10⁻³–10⁻¹	3–300	Shell; good filter sand
SP	10⁻³–10⁻¹	3–300	Shell; uniform — check filter compatibility both ways
SM	10⁻⁶–10⁻⁴	0.003–0.3	Semi-pervious; the classic internal-erosion problem soil
SC	10⁻⁸–10⁻⁶	3×10⁻⁵–0.003	Core material
ML	10⁻⁶–10⁻⁴	0.003–0.3	Erodible and dispersive-prone; avoid unfiltered
CL	10⁻⁹–10⁻⁷	3×10⁻⁶–3×10⁻⁴	Standard impervious core
MH	10⁻⁸–10⁻⁶	3×10⁻⁵–0.003	Poor fill; high compressibility
CH	10⁻¹⁰–10⁻⁸	3×10⁻⁷–3×10⁻⁵	Very low k, but shrink/swell cracking can dominate the field value

Ranges are for soil compacted near standard Proctor optimum. Compacting wet of optimum can drop k by one to two orders of magnitude versus the same soil compacted dry of optimum — the placement water content is often a bigger lever on k than the material selection.

Estimating k from Gradation

Hazen: k (cm/s) = C · D₁₀² (D₁₀ in mm, C ≈ 0.4–1.2, use 1.0)
Kozeny–Carman: k ∝ e³ / (1 + e) (void-ratio dependence within one soil)

Hazen is valid only for clean, loose to medium-dense sand with C_u < 5 and D₁₀ between about 0.1 and 3.0 mm. Outside that window it is not conservative in either direction. Treat it as a sanity check on a measured value, never as a substitute for one.

Test Methods & Their Valid Range

Method	Usable k (cm/s)	Notes
Constant head, rigid wall — ASTM D2434	> 10⁻³	Coarse soils; watch sidewall leakage and turbulence at high gradients
Falling head, rigid wall	10⁻³–10⁻⁶	Fine sands and silts
Flexible wall (triaxial) — ASTM D5084	≤ 10⁻⁶	The standard for liners and cores; back-pressure saturate first
Oedometer, from consolidation — ASTM D2435	10⁻⁷–10⁻¹⁰	Indirect: k = c_v · m_v · γ_w
Pumping test (field, saturated)	> 10⁻⁵	Best mass value; averages fabric and stratification over a large volume
Slug / bail test (field)	10⁻²–10⁻⁷	Samples only the material near the well screen
Double-ring infiltrometer — ASTM D3385	> 10⁻⁵	Vadose-zone infiltration rate, not saturated k; the usual BMP test
Borehole / Guelph permeameter	10⁻³–10⁻⁶	Field-saturated k above the water table

Field mass permeability commonly comes out 10 to 1,000× higher than a lab value on an intact tube sample, because the lab specimen misses the sand seams, root holes, desiccation cracks and lift interfaces that carry most of the flow. Use lab values for the core spec and field values for the seepage estimate.

Anisotropy & Temperature

Condition	k_h / k_v
Homogeneous, isotropic (an assumption, rarely a fact)	1
Compacted embankment fill placed in lifts	4–9
Natural stratified alluvium	2–10
Interbedded sand and clay, varved clay	10–100+

k₂₀ = k_T · (μ_T / μ₂₀)

Permeability is reported at 20°C. Because k scales inversely with viscosity, cold water moves roughly 40% slower at 5°C than at 20°C — a real effect for winter infiltration performance and for lab tests run in an unheated shed. Viscosity values are on the water properties card.

NRCS Hydrologic Soil Groups (K_sat of the least transmissive layer)

HSG	K_sat (in/hr)	K_sat (µm/s)	Typical texture
A	> 5.67	> 40	Sand, loamy sand, sandy loam — deep, well drained
B	1.42–5.67	10–40	Silt loam, loam
C	0.14–1.42	1.0–10	Sandy clay loam
D	< 0.14	< 1.0	Clay loam, silty clay, clay; also any soil over shallow bedrock or a high water table

The conductivity band applies when depth to a water-impermeable layer exceeds about 40 in. and depth to the seasonal high water table exceeds about 24 in. Shallower conditions force a dual group — A/D, B/D, C/D — which reverts to D unless the site is actually drained. These groups feed straight into the curve number card.

Two places k gets used badly. (1) Infiltration BMPs. A design rate is a field-measured rate divided by a factor of safety — commonly 2, and more where a single test represents a large footprint. Below about 0.5 in/hr, an infiltration practice is usually the wrong practice, not a practice with a longer drawdown. (2) Seepage and piping. Total seepage quantity scales with k, but internal erosion risk does not — it is governed by the exit gradient and by filter compatibility. The critical gradient for heave is i_c = (G_s − 1)/(1 + e) ≈ 1.0, and design exit gradients for embankment dams are held well below that (FS of 3 or more, i.e. i_exit ≤ ~0.3).

Sources: Terzaghi, Peck & Mesri, Soil Mechanics in Engineering Practice; Casagrande & Fadum permeability chart; NAVFAC DM 7.01, Soil Mechanics; USBR Design of Small Dams; ASTM D2434, D5084, D3385; NRCS National Engineering Handbook Part 630, Chapter 7 (hydrologic soil groups). Ranges are representative for preliminary work — site-specific testing governs final design.

Seepage through a dam? Open the embankment seepage calculator → · Piping check? Exit gradient · Filter criteria.

Related cheat sheets and tools

Get the group symbol first from the USCS classification card and the grain-size distribution tool — D₁₀ from that curve is the input to Hazen's formula. Viscosity for the temperature correction is on the water properties card; hydrologic soil group feeds the curve number card. Then run embankment seepage, exit gradient or filter criteria. For infiltration BMPs inside a full watershed model, see HydroComplete.